Sow body temperature and behavioral changes associated with the onset of estrus and ovulation

Willard, Naomi Christine

Permalink

https://hdl.handle.net/2142/127202 Copy

Description

Title

Sow body temperature and behavioral changes associated with the onset of estrus and ovulation

Author(s)

Willard, Naomi Christine

Issue Date

2024-11-22

Director of Research (if dissertation) or Advisor (if thesis)

Ellis, Mike

Doctoral Committee Chair(s)

Ellis, Mike

Committee Member(s)

• Knox, Robert
• Rodriguez Zas, Sandra
• Shull, Caleb
• Vande Pol, Katherine

Department of Study

Animal Sciences

Discipline

Animal Sciences

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

Sow body temperature, sow posture, estrus, ovulation

Abstract

The automated detection of the start of estrus and/or of ovulation could improve sow performance and reduce labor requirements. The objective of the research presented in this dissertation was to evaluate the potential for post-weaning changes in sow body temperature and/or behavior to predict the onset of estrus or ovulation. Two studies were carried out using similar designs and methodology. Study 1 used 99 sows and Study 2 used of 94 sows that were individually housed in gestation crates from weaning. Study 1 was carried out for the first 10 days post-weaning; Study 2 from day 2 to 7 post-weaning. Post-weaning changes in sow body temperature associated with the onset of estrus were evaluated in both studies; Study 2 also evaluated these changes associated with the onset of ovulation. Sows were checked for the start of estrus in the presence of a boar once per day from day 3 to 7 post-weaning. Ovulation was monitored using trans-rectal ultrasound which was carried out every 8 hours from the start of estrus. Sow rectal and body surface (on the ear, head, ham, and vulva) temperatures were measured every 2 hours between 0615 hours and 2300 hours. Post-weaning changes in sow posture associated with the onset of estrus were also evaluated in Study 1. Continuous video recording of the sows was carried out and the posture of each sow (lying, sitting, or standing) was recorded every 5 min. The percentage of observations that sows were active (i.e., sitting + standing) was calculated. Logistic regression analyses were carried out to develop equations to predict the onset of estrus or ovulation, separate analyses were carried out for each of the 3 days that sows first exhibited standing estrus (day 4, 5, and 6 post-weaning) or ovulated (days 5, 6, and 7 post-weaning). Equations to predict the onset of estrus based on changes in temperature used the data from Study 1 as the training dataset to develop the prediction equations and Study 2 as the testing dataset to validate the equations and vice versa. For the development of equations to predict the onset of estrus from posture measurements (Study 1) and of ovulation from temperature measurements (Study 2), the datasets were randomly divided into 2 populations: training dataset (70% of sows) and testing datasets (30% of sows). The dependent variable for the prediction of estrus was if the sow was in estrus or not and for the prediction of ovulation was if the sow had started ovulating or not. The independent variables for the prediction of estrus included changes in daily means for active postures (relative to day 2 post-weaning), changes in vulva temperature (between consecutive days), and the day post-weaning. The independent variables for the prediction of the onset of ovulation included changes in vulva temperature (between consecutive days), and the day post-weaning. For all analyses for the prediction of estrus, equations based only on day post-weaning generally had improved model accuracy statistics (AUROC, sensitivities, and specificity values, and error rates) compared to those based only on post-weaning body temperature changes. Equations to predict the start of estrus based only on changes in active postures had improved model accuracy statistics compared to those based only on changes in vulva temperature for all days sows were in estrus. However, equations based only on day post-weaning were generally more accurate at predicting the onset of estrus than those based on changes in active posture and/or vulva temperature. Similarly, equations to predict the onset of ovulation based on only day post-weaning generally resulted in improved model accuracy statistics compared to those based only on changes in vulva temperature. These results suggest that changes in posture or body surface temperature, either individually or in combination, were not as accurate as using day post-weaning to predict the onset of estrus or ovulation.

Graduation Semester

2024-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/127202

Copyright and License Information

Copyright 2024 Naomi Willard

Owning Collections